How this instrument works
A box-and-whisker plot summarizes a data set's distribution using just a handful of landmark values instead of every individual point. The 'box' spans the middle 50% of the data, from the first quartile (Q1) to the third quartile (Q3), with a line inside it marking the median. 'Whiskers' extend out from the box toward the minimum and maximum, though typically only as far as a defined fence — points beyond that fence are usually plotted as individual dots and treated as potential outliers rather than folded into the whisker itself.
This instrument computes every value that plot needs: the minimum and maximum of the data, Q1 and Q3 (using the same linear-interpolation quartile method already used throughout this site's percentile-based calculators), the median, the interquartile range (IQR = Q3 − Q1, the width of the box), and the lower and upper fences at Q1 − 1.5×IQR and Q3 + 1.5×IQR — the standard Tukey convention for flagging a point as an outlier. Everything here is drawn from formulas this site already ships individually as its own five-number-summary, quartile, and upper/lower-fence calculators; this page simply bundles them under the box-plot framing, since that's how most people go looking for them when they specifically want to draw or read this kind of chart.
The fences aren't hard cutoffs for 'impossible' values — they're a conventional rule of thumb (1.5 times the IQR beyond the box) for flagging points worth a second look, whether that means a data-entry error, genuine natural variation, or a real anomaly worth investigating. A tighter or looser multiplier is sometimes used in specific fields, but 1.5×IQR is the default nearly every statistics course and software package uses.
- Enter your data set into Data set, separated by commas, spaces, or new lines — at least 4 numbers are needed to draw a meaningful box plot.
- Read Minimum and Maximum for the two extremes of the data.
- Read Q1 and Q3 for the box's left and right edges, and Median (line inside box) for the line drawn inside it.
- Read Interquartile range (box width) for the width of the box itself, Q3 minus Q1.
- Read Lower fence and Upper fence for the standard 1.5×IQR outlier boundaries — any data point beyond either fence is conventionally flagged as a potential outlier.
Worked example — eight test scores
A small class's eight quiz scores are 2, 4, 4, 4, 5, 5, 7, 9 (out of 10). Entering that data set, Minimum reads 2 and Maximum reads 9. Q1 (lower quartile) reads 4.0, Median (line inside box) reads 4.5, and Q3 (upper quartile / right edge of box) reads 5.5 — so the middle half of the class scored between 4 and 5.5, a fairly tight cluster despite the wider 2-to-9 overall range.
Interquartile range (box width) reads 1.5, computed as 5.5 − 4.0. The fences come out to Lower fence = 4.0 − 1.5×1.5 = 1.75 and Upper fence = 5.5 + 1.5×1.5 = 7.75. Every one of the eight scores falls between those two fences, so this particular data set has no flagged outliers even though the highest score of 9 sits well above the box itself — it's simply within the whisker's normal reach, not beyond the fence.
Questions
What's the difference between a whisker and a fence?
The fence is a computed boundary (Q1 − 1.5×IQR on the low end, Q3 + 1.5×IQR on the high end); the whisker is the line actually drawn on the chart, which typically extends only to the most extreme data point that still falls within the fences, not all the way to the fence itself if no data point reaches that far. Points beyond the fence are usually plotted separately as individual dots rather than absorbed into the whisker.
Does a point beyond the fence mean it's definitely wrong or invalid?
No — the 1.5×IQR fence is a conventional flag for 'worth a closer look,' not proof of an error. A flagged point might be a data-entry mistake, but it might just as easily be genuine, unusual-but-real variation. Treat it as a prompt to investigate that specific observation further, not as an automatic reason to delete it.
Why does this calculator need at least 4 data points?
With fewer than 4 values there isn't enough data to meaningfully separate a lower quartile from an upper quartile — the box would collapse or become undefined in a way that doesn't represent a real distribution. Four is the practical minimum for the five-number summary (min, Q1, median, Q3, max) to say anything beyond just restating the raw data.
Is this the same information as this site's five-number-summary calculator?
It overlaps heavily by design — min, Q1, median, Q3, and max are the same five landmark values either way, since a box plot is essentially the five-number summary drawn as a chart. This page adds the IQR and the two 1.5×IQR fences on top, which are specifically the extra figures needed to draw the whiskers and flag outliers, framed for anyone searching specifically for 'box plot' numbers rather than the summary alone.
Why is 1.5 the standard multiplier for the fences instead of some other number?
It's John Tukey's original convention from when he introduced the box plot, chosen because it flags a reasonably small, useful fraction of points as worth reviewing for roughly bell-shaped data without being triggered by ordinary variation. Some fields use a stricter 3×IQR fence for flagging only 'extreme' outliers, but 1.5×IQR is the default nearly universal across statistics courses, software, and this instrument.